Spot

Inspiration

• We love dogs and want to share it w the world (Twitter)!!
• This can definitely be extended to other applications, but we decided to narrow it to dogs

What it does

• Display image of a dog (or literally anything) to the user,
• Interpret user's facial reaction of the image using sentiment analysis,
• Calculate user's rating of the image,
• Post the rating and the image on the user's Twitter account (@susanscfae)

How I built it

Twitter API Tensorflow QT Principal Component Analysis MATLAB

Challenges I ran into

• Data collection was difficult in a noisy data collection because of the time of night at which the data collection was conducted
• Using a loaned laptop from the MLH hardware lab meant that there were a lot of development tormentor tools that had to be set up: GitHub, python, MATLAB, tensor-flow. Overall these challenges were identified and solved in a timely manner

Accomplishments that I'm proud of

• GUI for this application
• General improvement in tester's mood after looking at our doggos
• Working with tensorflow

What I learned

• Using tensorflow
• Using PyQt

What's next for Spot

Using this for other applications. Here are some of our ideas:

• How a person's initial reaction to a product influences their purchase decision
• Gauge general confidence in the financial market
• Assessment of a general wellbeing
• Predicting how well group/teams of people will work together
• Focus groups for a product

Built With

• python
• twitter-ads-api
• qt
• opencv